In-Situ Diagnostic Methods for SOFC G. Schiller, K.A. Friedrich, M. - - PowerPoint PPT Presentation

In-Situ Diagnostic Methods for SOFC

• G. Schiller, K.A. Friedrich, M. Lang, P. Metzger, N. Wagner

German Aerospace Center (DLR), Institute of Technical Thermodynamics, Pfaffenwaldring 38-48, D-70569 Stuttgart, Germany

International Symposium on Diagnostic Tools for Fuel Cell Technologies, Trondheim, Norway, June 23-24, 2009

Outline

Introduction Electrochemical Impedance Spectroscopy on Stacks Spatially Resolved Measurements: Current density Voltage Impedance Temperature Gas Composition Optical Spectroscopy X-Ray Tomography Conclusion

Investigation of Degradation and Cell Failures

Insufficient understanding of cell degradation and cell failures in SOFC Extensive experimental experience is not generally available which would allow accurate analysis and improvements Long term experiments are demanding and expensive Only few tools and diagnostic methods available for developers due to the restrictions of the elevated temperatures

Conventional Test Stand Diagnostics

Conventional test stand diagnostics: provide important and essential information about fuel cell performance and behaviour: U(i) characteristics, OCV… EIS on single cells Current interrupt methods Performance degradation with time U(t); i(t) … Cell voltage distribution Ustack = U1 + U2 + U3 …. Pressure loss / Gas tighness test Gas utilization measurement Temperature distribution and control

„Sophisticated“ (non-traditional) in-situ Diagnostics

Electrochemical impedance spectroscopy on stacks Spatially resolved measuring techniques for current, voltage, temperature and gas composition Optical imaging Optical spectroscopy Acoustic emission detection X-ray tomography

Challenges for EIS for Stack Investigations

Large areas (e.g. 100 cm2) lead to high current and low impedances of about 1 mOhm. Electrochemical processes appear at high frequencies (up to 100 kHz) due to the high reaction rates at high temperatures. Stacks generally contain metallic components leading to high frequency disturbances. Contacting of all cells and sensing in specific cells does not account for the voltage distribution in the stack. The sensor wires are at high temperatures: an optimization of the measurement system is not possible during operation. Strong overlapping of electrode processes; evaluation with equivalent circuits can be inaccurate. For system with current > 40 A no commercial equipment available.

Mitigation

• f EIS Problems

Reduction of the high frequency disturbances by optimization of the wiring of the electrical sensing of the SOFC stack. Variation of the operating conditions (gases, temperature) in

• rder to determine the different impedances of the electrode

processes Modeling of the spectra by an equivalent circuit. Development of advanced EIS equipment for high currents / high frequencies in corporation with instrument manufacturer (Zahner Elektrik GmbH).

Experimental Set-up for EIS Measurements

• f Stacks

at DLR

CSZ05-DGF09-CT, 750°C 5H2+5N2+3%H2O / 20air (SLPM) 94h

0,0 0,2 0,4 0,6 0,8 1,0 1,2 200 400 600 800 current density i [mA/cm²] cell voltage U [V] 200 400 600 power density p [mW/cm²]

cell 5 (top) cell 4 cell 3 cell 2 cell 1 (bottom)

p U

Performance of the 5-Cell Short Stack at 750°C

(5 H2 +5 N2 +3%H2 O / 20 air (SLPM), 94 h) Cell 1-4 : 1.10 V Cell 5 : 1.05 V

Cell 5: 404 mW/cm² Cell 4: 476 mW/cm² Cell 3: 447 mW/cm² Cell 2: 472 mW/cm² Cell 1: 415 mW/cm² @ 3,5V Pstack = 184 W FU = 37%

• 0,5
• 0,25

0,25 0,5 0,75 1 1,25 0,25 0,5 0,75 1 1,25 1,5 1,75 2 2,25

Re Z [Ohm*cm²] Im Z [Ohmcm²]

0 mA/cm2 60 mA/cm2 120 mA/cm2 180 mA/cm2 240 mA/cm2 300 mA/cm2 360 mA/cm2 420 mA/cm2

80 Hz

5-Zellen Short Stack [CSZ-05-83-CT], cell 5 T=750°C, Zellfl. 84cm²

Gas Concentration Cathode Anode

50mHz 1 Hz 6 Hz 100 kHz

Nyquist Plot of one Cell of a 5-Cell Short Stack at Different Current Densities (750°C, 2.5 H2 +2.5 N2 / 20 air (SLPM), 142 h) 0.14 - 0.17 cm2 0.55 - 2.2 cm2

Equivalent Circuit for the Fitting of the Impedance Spectra

Gas Concentration Inductivity Cathode Ohmic R Rp (C) ZL Anode Cdl (A) RP (A) Cdl (C) RN (A) CN (A)

CSZ-05-83-CT, 750°C 2,5H2+2,5N2 / 21 Air (SLPM) cell 5, 142 h

0,0 0,2 0,4 0,6 0,8 1,0 1,2 50 100 150 200 250 300 350 400 450 current density i [mA/cm²] cell voltage U [V] 20 40 60 80 100 voltage loss [mV]

U cell ΔU (Anode) ΔU (Cathode) ΔU (Gas Concentr.) ΔU (Ohm) Cell5 @ 700mV : 380 mW/cm2 U cell ΔU

Voltage Losses at one Cell of a 5-cell Short Stack at Different Current Densities (750°C, 2.5 H2 +2.5 N2 / 20 air (SLPM), 142 h)

Contact resistances Polarisation resistances

Motivation

Strong local variation of gas composition, temperature, current density Distribution of electrical and chemical potential dependent on local concentrations of reactants and products Reduced efficiency Temperature gradients Thermo mechanical stress Degradation of electrodes

H2 H O

2

O2 O2 H2 H O

2

O2

Measurement Setup for Segmented Cells

16 galvanically isolated segments Local and global i-V characteristics Local and global impedance measurements Local temperature measurements Local fuel concentrations Flexible design: substrate-, anode-, and electrolyte-supported cells Co- and counter-flow

Cell design and Testing Station

From a „simple“ cell design with manually controlled features GC measurement Flexible housing, impedance spectra with reduced interferences Assembly and contacts All cell concepts Improved contacting Reliable assembly Impedance measurement Temperature measurement

Schematic Lay-out of the Electrical Circuit of the Segmented Cell Configuration

current busbar equipotential line current busbar equipotential line

Internal cell resistances: Ri,j, Resistances of the wires contacting the anode: RLA,j Resistances of the wires contacting the cathode: RLK,j Only segments 1, 2, 3, 16 are illustrated

OCV Voltage Measurement for Determination of Humidity

• Voltage distribution at standard flow rates:
• 48.5% H2

, 48.5% N2 + 3% H2 O, 0.08 SlpM/cm² air

13 14 15 16 9 10 11 12 5 6 7 8 1 2 3 4

fuel gas air

         

2 2 2

ln

H O O H rev rev

p p p zF RT U U

Nernst equation: Produced water: S4: 0.61%, S8: 0.72%, S12: 0.78%, S16: 3.30%

Variation of Load

• Reformate

Anode supported cell, LSCF cathode, 73,96 cm², gas concentrations (current density equivalent): 54.9% N2 , 16.7% H2 , 16.5% CO, 6,6% CH4 , 2.2% CO2 , 3.2% H2 O (0.552 A/cm²), 0.02 SlpM/cm² air

0,0 50,0 100,0 150,0 200,0 250,0 300,0 Segment 9 Segment 10 Segment 11 Segment 12

power density p [mW/cm²]

0,0 15,0 30,0 45,0 60,0 75,0 90,0

fuel utilisation fu [%]

p(i) 100 mA/cm² p(i) 200 mA/cm² p(i) 400 mA/cm² p(i) 435 mA/cm² fu 100 mA/cm² fu 200 mA/cm² fu 400 mA/cm² fu 435 mA/cm²

fu

100 200 400 435 100 200 400 435

Power density mW/cm2 Fuel utilisation (%)

0,05 0,1 0,15 0,2 0,25 0,3 Segment 9 Segment 10 Segment 11 Segment 12

Gaskonzentration / %

H2 CO CH4 CO2 H2O KS4X050609-7 in Metallischem Gehäuse; Substrat: Anodensubstrat, aktive Zellfläche:73,78 cm²,A: 542 µm NiO/YSZ, E: 14 µm YSZ + YDC, K: 28 µm LSCF, Kontaktierung: 30 µm LSP16+Pt3600, Integral, Gasflüsse: 0,552 A/cm² Stromdichteäquivalent (54,9% N2, 16,7% H2, 16,5% CO, 6,6%CH4, 2,2%CO2, 3,2% H20) // 0,08 SlpM/cm² Luft, 800 °C, 0 mA/cm²

Reformate: Changes

• f the

Gas Composition at 0 mA/cm²

Metallic housing, anode substrate, active area 73.78 cm² Anode: 542 µm NiO/YSZ, Electrolyte: 14 µm YSZ + YDC, Cathode: 28 µm LSCF Operation conditions: 0.10 A/cm² - Anode  = 5.52 (54.9% N2 , 16.7% H2 , 16.5% CO, 6.6% CH4 , 2.2% CO2 , 3.2% H2 O 0.08 Nlpm/cm² Air, 800°C)

Concentration / %

9 10 11 12 H2 CO H2 O CO2 CH4

Alteration of the gas composition at 435 mA/cm²

0,05 0,1 0,15 0,2 0,25 0,3 Segment 9 Segment 10 Segment 11 Segment 12

Gaskonzentration / %

H2 CO CH4 CO2 H2O

Concentration / %

H2 10 11 12 CO CH4 H2 O CO2 9

Combined Experimental and Modeling Approach

Objectives of the study: Better understanding of the local variations Identification of critical conditions Optimisation of cell components Experiments on single Experiments on single segmented SOFC segmented SOFC Electrochemical model of Electrochemical model of local distributions local distributions

H2 H2/CO CH4 H2O CO2 anode electrolyte cathode O2/N2 N2 x y x y elyt elde gas z interconnector interconnector

Potential for Optical Spectroscopies

Raman spectroscopy Laser Doppler Anemometry (LDA) Particle Image Velocimetry (PIV) Fast-Fourier Infrared (FTIR) Coherent Anti-Stokes Raman Spectroscopy (CARS) Electronic Speckle Pattern Interferometry (ESPI)

Digital CCD camera Distance microscope (resolution1 µm) Quarz window Transparent flow field Imaging spectrograph Lenses/filter Pulsed Nd:YAG laser (532 nm, 10 ns) Open tube (5 mm) a) In situ microscopy b) In situ Raman laser diagnostics

15 cm

Heat & radiation shield SOFC

neutron tomography in-situ synchrotron radiography in-situ neutron radiography

Tomography Diagnosis of PEM Fuel Cells

Investigation of water management under operating conditions

X-Ray Tomography (CT) Facility at DLR

3 dimensional non intrusive imaging of SOFC cassette X-Ray CT Facility v|tome|x L450 at DLR Stuttgart

Summary

The operating conditions (elevated temperature) reduce significantly the possibilities for in-situ SOFC diagnostic methods. EIS will remain the main diagnostic probe of the state of SOFC. Non-traditional in-situ diagnostics methods can provide additional important information: Spatially resolved measurements to obtain local distribution

• f cell properties (current, voltage, impedance, gas compo-

sition, temperature) Combined analytical and modeling approach Large future potential for optical spectroscopies (e.g. Raman spectroscopy) and x-ray tomography.